LateralGM is designed to work back to Java 1.5 (although 1.5 requires an additional jar). If a person wants to use 1.5 or 1.6 for whatever reason, we will support them as best we can. Some day we may stop supporting older versions, but for now, we support them.

As for gra's problem, actually I followed the exception stack trace that he provided, and it indicated that he was using an older version, as it pointed to code that has since been corrected. If he can reproduce the problem and have the stack trace indicate that it is the latest version, I'll believe it.

While I certainly don't oppose the idea, I do, however, feel like this is just another case of "stick it on github and the problem is magically solved". The problem with this ideology is twofold: 1) Interested manpower. Most of our interested manpower is already working on the wiki and enigma, which have higher priority than the bot and the website (as they should). As for generating interest, that leads to point 2) Generalizability. The bot and the website are pretty much singletons. They don't generalize to anything else, so there's not really a need for a second one to be launched anywhere else on the entire internet. Generally a project is best suited for source distribution and version control when it's expected that there will be multiple instances of it.

This leads to a more serious issue: more leechers than seeders. With so few people actually contributing to these projects, the only thing opening the source code effectively does is allow people to read it, learn from it, and abuse it. We're revealing our security flaws, but are we fixing them? If we're not, people will use those security flaws.

Just my two cents and concern. At any rate, I'd love to see them both worked on more, but I think we should hesitate a little before we just jump to "throw it on github for all to see".

Huh, I don't recall exactly, but maybe the modular resource system was implemented only in the joshedit branch at the time. Namely, this changed a way to access certain variables (like gmObjects) from static to a map lookup. This caused a serious incompatability with LGM and the plugin, and since this was only implemented in JoshEdit, it means the plugin would not be able to switch between master and joshedit branches (which is exactly what Robert did), hence the error. The quickfix is to use a different version of the plugin.
I think the ideal fix is to ensure that the latest official LGM and the latest official plugin are built, as they have mostly stabilized at this point and should be compatible again. If you're still having issues after that, let me know, as you might have found a regression.

You could fudge it and do a sleep(1) every few frames.

This problem is fixed, but you still have an outdated jar. Robert's build must have failed or something.

Room speed can be set to 0 as of f2c5c2f89adc. Robert has been maintaining the latest builds, so he'll probably build it for you.
Note that this behavior may lead to incompatibilities in GM.

gra: I think I found the source of the problem, but I'm not sure if it's the same as your problem. Please check out my comments on the ticket and provide additional info.

This sounds like a LateralGM bug, so it belongs on the LateralGM tracker, not the Enigma tracker (the link at the top). But it looks like gra's going the correct route anyways.

make[1]: i486-mingw32-g++: Command not found

Seems to suggest you don't have g++ installed? Or maybe it's expecting a specific version of g++ which isn't yours.

Regardless of packages, those are essential libraries that must be obtained in whatever way mac provides, and I'm fairly certain the mac installation instructions and dependencies pages on the wiki explain how to get those sufficiently.

Also, if anything broke, it's Robert's fault.

LateralGM is proud to state that these changes have been merged into the official LateralGM repository with appropriate revision to make them acceptable.

Robert has expressed that he does not desire to continue to maintain a separate fork of LateralGM, and his changes were always intended to be merged back into the main LateralGM repository maintained by myself.

Note that for the time being, Robert will be in charge of understand exactly what this implies for persons wishing to checkout or build from the LateralGM repository (namely, which branch to use), at least until I can figure it out for myself.

Technically I think this was one of the missing features of JoshEdit that was holding me back from merging.

The Override tag will not directly cause any implications. They simply prevent the programmer from doing something stupid like mistyping an override function (capitalization, for example) and ending up with inherited behavior when you thought you overrode it. They're just a preventitive measure, so to remove them will not break anything.

I just wanted to clarify that this "LGM 1.8" is Robert's fork of LGM, done with my permission.
Robert's fork uses a different version numbering system than the official repository (official is still on 1.6 and may be for some time), and also may have some minor licensing conflicts, regressions, or unapproved features that will (hopefully) not appear in the official LateralGM repository and releases maintained by myself.

That said, I'm glad to see he's managed to make his own fork release, and glad to see such a positive receiption for his release.